Characteristics of atmospheric-pressure, radio-frequency glow discharges operated with argon added ethanol

Transcription

1 JOURNAL OF APPLIED PHYSICS 101, Characteristics of atmospheric-pressure, radio-frequency glow discharges operated with argon added ethanol Wen-Ting Sun, Guo Li, He-Ping Li, a Cheng-Yu Bao, Hua-Bo Wang, b and Shi Zeng Department of Engineering Physics, Tsinghua University, Beijing , People s Republic of China Xing Gao School of Public Health and Family Medicine, Capital University of Medical Sciences, Beijing , People s Republic of China Hui-Ying Luo Beijing Center for Diseases Control and Prevention, Beijing , People s Republic of China Received 4 December 2006; accepted 4 May 2007; published online 20 June 2007 Rf, atmospheric-pressure glow discharge APGD plasmas with bare metal electrodes have promising prospects in the fields of plasma-aided etching, thin film deposition, disinfection and sterilization, etc. In this paper, the discharge characteristics are presented for the rf APGD plasmas generated with pure argon or argon-ethanol mixture as the plasma-forming gas and using water-cooled, bare copper electrodes. The experimental results show that the breakdown voltage can be reduced significantly when a small amount of ethanol is added into argon, probably due to the fact that the Penning ionization process is involved, and a pure -mode discharge can be produced more easily with the help of ethanol. The uniformity of the rf APGDs of pure argon or argon-ethanol mixtures using bare metallic electrodes is identified with the aid of the intensified charge coupled device images American Institute of Physics. DOI: / I. INTRODUCTION Although low-pressure plasmas have found their wide applications in the last few decades, operating at reduced pressures requires expensive and complicated vacuum system which results in the high capital costs, the size limitations on the treated objects, the complex robotic assemblies used to shuttle materials in and out of vacuum chamber, etc. 1 In recent years, different kinds of atmospheric-pressure nonequilibrium discharges APNEDs, such as the dielectric barrier discharge DBD plasma, 2 the plasma needle, 3 the cold arc-plasma jet, 4 the one atmosphere uniform glow discharge plasma OAUGDP, 5 the surface-wave discharge, 6 the microhollow cathode discharge, 7 the radio-frequency atmospheric-pressure plasma jet, 8,9 etc., are often covered under the rather broadly used term atmospheric-pressure glow discharge APGD. But some of them cannot be directly connected to the standard glow discharges. The common feature in all these discharge plasma sources is that they attempt to produce nonequilibrium low temperature plasmas at atmospheric pressures. So, in this paper, the term glow discharge symbolizes nonequilibrium plasma as a general terminology. Among different kinds of APGD plasma sources, the APGD plasmas using bare metal electrodes driven by rf power supply developed in recent years have attracted much attention of the researchers Besides the removal of the vacuum chamber, the breakdown voltage of the rf APGD plasmas with bare metal electrodes can be reduced significantly compared with atmospheric-pressure a Author to whom correspondence should be addressed; FAX: ; electronic mail: liheping@tsinghua.edu.cn b Present address: Infinova Shen Zhen Ltd., Shen Zhen , Guangdong Province, People s Republic of China. DBDs resulting from the elimination of the dielectric s covered on the electrodes or placed between electrodes in DBDs. 1 With the foregoing outstanding features, the rf APGD plasmas using bare metal electrodes have shown bright prospects to potentially replace low-pressure discharge devices in some existing applications and to create other applications in future, such as plasma-aided etching in microelectronic industry, 16,17 plasma-enhanced chemical vapor deposition of silicon nitride or silicon dioxide films, decontamination of chemical and biological warfare CBW agents, 21 decommissioning of radioactive and chemical waste, 22 inactivation of micro-organisms, 23,24 graffiti removal, car wash, 25 and so on. Usually, in the rf APGD plasmas using bare metallic electrodes, the primary working gas is helium, into which a small fraction 0.5% 3% of reactive gases e.g., O 2,CF 4, water vapor, etc. may be added in order to generate a flux of chemically active species. 8,16 23,26 29 Although it is easier to obtain a uniform glow discharge with pure helium as the primary plasma-forming gas, the consumption of expensive helium increases the operation costs for this case. Recently, studies on the characteristics of rf APGD plasmas are reported with pure argon 30 or argon-oxygen 1% in volume 31 as the plasma-forming gas. Although the operation costs can be reduced using argon, instead of helium, the breakdown voltage for argon is much higher than that for helium. 26,30 As is well known, the higher breakdown voltage may cause a rapid multiplication of electrons after breakdown and lead to the formation of streamers or a filamentary arc. 26 In Ref. 32, it was also pointed out that it was difficult to initiate and maintain the stable, atmospheric-pressure rf -mode glow discharge plasmas with argon due to the high breakdown /2007/ /123302/6/$ , American Institute of Physics

2 Sun et al. J. Appl. Phys. 101, FIG. 1. Schematic diagram of the experimental setup a, the coaxial-type plasma generator b, and the planar-type plasma generator c. voltage. Therefore, it is necessary to study the discharge mechanisms of the rf glowlike discharge plasmas using bare metal electrodes under different operation conditions in order to achieve stable, uniform glowlike discharges using a variety of plasma working gases and at lower breakdown voltages. In this study, the discharge characteristics of rf APGDs, including the breakdown voltage of the gas, the relationship between the discharge mode after breakdown and the gap spacing, etc., are investigated with pure argon and argonethanol mixture as the plasma-forming gas and using bare metal electrodes. II. EXPERIMENTAL SETUP A schematic diagram of the experimental setup is shown in Fig. 1 a. Two types of plasma generators are employed in this experiment as shown in Figs. 1 b and 1 c, respectively. The coaxial-type plasma generator, as shown in Fig. 1 b, consists of two 95 mm long, coaxial, water-cooled copper electrodes. The inner diameter of the outer electrode, which is employed as the grounded electrode, is 19.2 mm, while the central electrode is rf MHz powered with a diameter of 16.0 mm. Figure 1 c shows the planar-type plasma generator which is composed of two 5 8 cm 2 planar, bare, water-cooled copper electrodes, i.e., the rf powered top electrode and the grounded bottom electrode. Teflon spacers are used to seal the plasma generator on both sides and adjust the distance between the electrodes. The powered electrodes of the plasma generators are connected to the rf power supply through a matching network. The plasma-forming gas argon, or argon-ethanol mixture with argon passing through a homemade chamber in which the argon is mixed with ethanol enters the discharge region, ionized between electrodes under the applied rf electric field, and flows out of the generator. The concentration of ethanol in the argon-ethanol mixture is adjusted by controlling the number of the graduated cylinders placed in the chamber. A pipette with accuracy of 0.02 ml is used to measure the total consumed ethanol V ethanol at a time interval t. If the argon flow rate is Q Ar, the ethanol concentration can be expressed as C ethanol = ethanol V ethanol / tq Ar mg/l, where ethanol is the mass density of the ethanol. The rms values of the voltage and current are measured using a high voltage probe Tektronix P5100 and a current probe Tektronix TCP202, respectively, and recorded on a digital oscilloscope Tektronix TDS3034B. The discharge pictures are taken by a digital camera Fujifilm S5500 and an intensified charge coupled device iccd Type DH734-18F-03/W/P43 camera, respectively. With the Andor istar 734 iccd, very short time acquisition down to 5 ns exposure time of the discharge can be performed in order to detect the presence of streamers in the discharge gap. By removing the cameras, the spectra of the discharge are measured by a monochromator WDG30, Beijing Optical Instrument Factory, China plus photomultuplier tube CRC131-01, Beijing Hamamatsu Photon Techniques Inc., China system with wavelengths ranging from 200 to 800 nm. In this paper, for investigating the relationship between the gas breakdown voltage and the gap spacing, and also for taking the pictures using the iccd conveniently, the planar-type plasma generator shown in Fig. 1 c is employed, while other measurements are conducted using the coaxial-type plasma generator shown in Fig. 1 b with constant gap spacing d=1.6 mm.

3 Sun et al. J. Appl. Phys. 101, FIG. 2. Color online I-V curves for pure argon and argon-ethanol rf APGDs using the coaxial-type plasma generator, Q Ar =5.0 slpm. III. EXPERIMENTAL RESULTS A. Electrical measurements Similar to the rf capacitive discharges at intermediate pressure, the rf APGD can assume two different but stable operation modes, i.e., mode, which is sustained by volumetric ionization processes, and mode, in which ionization by secondary electrons from electrode surfaces is important. 27,33 Secondary electron emission strongly influences the gas ionization in the mode discharge but matters little in the mode dishcarge. 10 The two modes differ in the intensity and luminosity distribution along the discharge length. When the mode appears, it contracts so that the current density at the electrodes rises and the discharge column changes its shape radically at each electrode. 33 In this experiment, due to the large area of the electrodes about 40 cm 2 for both the coaxial-type and planar-type plasma generators, the mode discharge can only cover a small part of the electrodes due to the limitation of the maximum power output of the rf power supply used in this lab. But the mode discharge can cover the whole gap space. In this paper, we distinguish the two discharge modes by visual difference, i.e., there is a bright negative glow very close to both electrodes with a few filamentlike contracted positive columns for argon mode discharge. The similar results were also reported by Laimer et al. 30 The measured current and voltage characteristics, i.e., the so-called I-V curves, of the discharge process for pure argon and argon-ethanol mixtures with different ethanol concentrations are shown in Fig. 2 for the case with the argon flow rate Q Ar =5.0 standard liters per minute slpm. Figure 2 shows that for the discharge process with pure argon, the ignition occurs at point A with a rather high breakdown voltage 566 V. After breakdown, the discharge voltage drops sharply to 134 V point B, and a uniform glow discharge operated in mode appears fully between electrodes with the power input of about 60 W. Then, with the increase of the discharge current B-C, the discharge voltage slightly increases, and the -mode discharge is maintained. With the continuous increase of the discharge current, a -mode discharge or arcing may occur after point C associated with larger rf power input. On the other hand, for the discharge processes with argon-ethanol mixtures, the ignitions occur with much lower breakdown voltages. The lowest breakdown voltage is 139 V for C ethanol =0.7 mg/l, while 166 and 170 V for C ethanol =0.2 and 4.7 mg/l, respectively, in FIG. 3. Color online Photographs of the discharge after breakdown using the coaxial-type plasma generator, Q Ar =5.0 slpm, a pure argon; b argonethanol mixture C ethanol =0.2 mg/l. this study. For all the cases with argon-ethanol mixture as the plasma-forming gas, the discharge voltages drop slightly after breakdown, and the uniform -mode discharge only partially covers the electrodes due to the low power input about 30 W. Then, with the increase of the discharge current, the -mode discharge region becomes larger and can cover all the electrode surfaces, accompanied with the increase of the discharge voltage. And similarly, as the discharge current is increased continuously, arcing or -mode discharge usually occurs associated with the larger power input. The measurements presented in Fig. 2 are all repeated three times for the discharge processes with argon or argon-ethanol mixtures, and the standard deviations of the measured discharge voltages are all smaller than 9 V. It can be seen from Fig. 2 that with the help of ethanol, the breakdown voltage drops significantly compared with the case of pure argon discharge, i.e., the lowest breakdown voltage with the ethanol concentration 0.7 mg/l is only 1/4 of that at point A corresponding to the discharge with pure argon. In Ref. 30, it was reported that an - coexisting mode always occurs after breakdown. It is found in this study that and/or mode can appear directly after breakdown, depending on the gap spacing between electrodes, the matching network of the circuit, etc. It is also found that the -mode discharge, instead of the - coexisting mode, can be achieved much more easily after breakdown for the argonethanol rf APGDs with bare metal electrodes. As seen in Fig. 3 a, for Q Ar =5.0 slpm and power input P in =113 W, a -mode discharge with several filamentlike columns coexists with a uniform -mode discharge after breakdown. However, for the same argon flow rate, a pure -mode discharge can be obtained after breakdown with the ethanol concentration 0.2 mg/l, as shown in Fig. 3 b at a much lower input power

4 Sun et al. J. Appl. Phys. 101, FIG. 4. Color online Relationship between the breakdown voltage and the gap spacing with pure argon and argon-ethanol mixture using the planartype plasma generator, Q Ar =5.0 slpm. P in =28 W. And with increasing the input power, the pure -mode discharge can cover the full space between electrodes. In order to study the influence of the gap spacing d on the breakdown voltage V b and the discharge mode after breakdown, the planar-type plasma generator, as shown in Fig. 1 c, is employed in this study. Figure 4 shows that with constant argon flow rate Q Ar =5.0 slpm 1 for both cases with pure argon or argon-ethanol mixture as the plasmaforming gas, the breakdown voltages increase with increasing the gap spacing between electrodes; 2 at the fixed gap spacing, the breakdown voltage for pure argon discharge is much higher than that for the case of argon-ethanol discharge; 3 for argon-ethanol discharge, the critical value of the gap spacing between electrodes for obtaining a pure -mode discharge after breakdown is 2.5 mm, which is much larger than the corresponding value for the case of pure argon discharge 1.3 mm. When the gap spacing exceeds its critical value, a -mode or - coexisting mode discharge may occur after breakdown. FIG. 5. Variations of the breakdown voltage with different ethanol concentrations using the coaxial-type plasma generator, Q Ar =5.0 slpm. The relationship between the breakdown voltage V b and the ethanol concentration C ethanol is also studied with the coaxial-type plasma generator in this paper. It is found that an optimum value of the ethanol concentration exists corresponding to the lowest breakdown voltage. As shown in Fig. 5, when the ethanol concentration is about 0.7 mg/l, the lowest breakdown voltage appears in this study, while both lower and higher ethanol concentrations than this optimum value result in higher breakdown voltages. The similar phenomena are also observed by using the planar-type plasma generator. B. Discharge images To identify the uniformity of the discharges with argon and argon-ethanol mixtures, the discharge images are taken in a short exposure time T ex =10or30ns with the Andor istar 734 intensified CCD camera, and also compared with the pictures taken by the digital camera Fujifilm S5500. In Fig. 6, the pictures listed on the left column, i.e., Figs. 6 a 6 d, are taken by the iccd with a short exposure time FIG. 6. Color online Discharge images taken by the iccd camera a d and the digital camera e h using the planar-type plasma generator, Q Ar =5.0 slpm, d=1.24 mm a -mode discharge of pure argon, T ex =10 ns, P in =65 W; b -mode discharge of pure argon, T ex =10 ns, P in =150 W; c -mode discharge of argon-ethanol mixture 0.2 mg/l, T ex =10 ns, P in =80 W; d -mode discharge of argon-ethanol mixture 9.5 mg/l, T ex =30 ns, P in =20 W; e -mode discharge of pure argon, T ex =2.5 ms, P in =65 W; f -mode discharge of pure argon, T ex =2.5 ms, P in =150 W; g -mode discharge of argonethanol mixture 0.2 mg/l, T ex =2.5 ms, P in =80 W; h -mode discharge of argon-ethanol mixture 9.5 mg/l, T ex =125 ms, P in =20 W.

5 Sun et al. J. Appl. Phys. 101, FIG. 7. Color online Emission spectra of the excited argon and oxygen atoms with different ethanol concentrations, Q Ar =5.0 slpm, d=1.24 mm. T ex =10 ns for Figs. 6 a 6 c and 30 ns for Fig. 6 d, while their counterparts on the right column, i.e., Figs. 6 e 6 h, are taken by the digital camera with a longer exposure time T ex =2.5 ms for Figs. 6 e 6 g and 125 ms for Fig. 6 h. In Figs. 6 d and 6 h, the longer exposure time is employed compared with their counterparts because of the low emission intensity of the discharge with high ethanol concentration 9.5 mg/l. Figure 6 clearly shows the laterally uniform rf APGDs without any streamers. C. Spectroscopic measurements and discussions In Ref. 34, the influence of the ethanol CH 3 CH 2 OH on the behaviors of the electron and ion densities and distribution of metastable Ar * in flowing afterglow plasmas were investigated in the pressure range of Pa. It was indicated that a comparatively sharp increase in both electron and ion densities occurred when the titration vapor ethyl alcohol was added into the plasma-forming gas argon due to the Penning ionization Ar * +CH 3 CH 2 OH B D 1 + e, 1 where B + 1 and D 1 represent the products of ions and neutrals, respectively, while e stands for the electrons. 34 It was also indicated in Ref. 35 that the bands between carbon and hydrogen atoms in the alcohols could be broken in an argon surface wave sustained discharge at atmospheric pressure. Because the internal energy of the metastable Ar * ev is higher than the ionization energy of an ethanol molecule ev, 34 it is expected that the Penning ionization 1 may also occur in the present argon-ethanol rf APGDs, which could lead to a great decrease of the breakdown voltage. The emission spectra of the discharges with different ethanol concentrations are performed by optical emission spectroscopy in this study. The excited argon atomic lines at 696.5, 707, 715, 727, 750, 763.5, and nm Refs. 31 and 36 are depicted in Fig. 7 for the rf APGD plasmas with argon as the primary plasma-working gas. The corresponding spectra of hydroxyl OH molecular band A 2 +, =0 X 2, =0, nm 37 under the same operation conditions are shown in Fig. 8. The OH molecular spectrum nm and the excited oxygen atom emission line FIG. 8. Color online Emission spectra of hydroxyl molecular OH band with different ethanol concentrations, Q Ar =5.0 slpm, d=1.24 mm 777 nm are also seen in Figs. 7 and 8 even for the pure argon discharge. One of the reasons for this phenomenon is that the discharges are generated in ambient air. 31,36 It can be seen from Figs. 7 and 8 that 1 the introduction of the ethanol is one of the possible reasons which results in the production of the radical of hydroxyl OH due to the Penning ionization process discussed above; 2 the intensity of the hydroxyl band increases sharply at first with a small concentration of ethanol 0.1 mg/l, and then, with the further increase of the ethanol concentration, the intensity of hydroxyl band decreases; 3 the emission intensities of the excited argon and oxygen atomic lines decrease with increasing the ethanol concentration. In Ref. 38, it was indicated that ionization through metastables was one of the important processes in increasing the plasma density; and the densities of the argon metastables Ar * and electrons, as well as the optical emission intensities, decreased with the increase of N 2 ratio in an argonnitrogen mixture at a low pressure range Torr due to the stepwise ionization and metastable pooling. In this study, the argon metastables Ar * may play the similar role as that discussed in Ref. 38 in the discharge processes for the argon-ethanol rf APGDs due to the Penning effect 1, since the emission intensities, including the optical emission intensity of argon metastables Ar * at nm, 38 of the argonethanol discharges are lower than those for a pure argon discharge, as shown in Fig. 7. Because of the very complex chemical processes involved in the argon-ethanol plasma system, the deeper investigations on the discharge mechanisms and chemical reactions in the argon or argon-ethanol rf APGDs need to be conducted in future work. IV. CONCLUSIONS In this paper, the discharge characteristics of the rf APGD plasmas generated with pure argon or argon-ethanol mixture as the plasma-forming gas and using water-cooled, bare copper electrodes are investigated. The experimental results show that a small fraction addition of ethanol into argon can significantly reduce the breakdown voltage of the rf APGD plasmas generated between the bare metal electrodes, and make it easier to obtain more uniform glowlike discharges operated in mode after breakdown compared with

6 Sun et al. J. Appl. Phys. 101, the pure argon discharge, probably due to the Penning ionization process involved in the argon-ethanol rf APGDs. The uniformity of the rf APGDs with the pure argon or argonethanol mixture using bare metal electrodes is also identified by the iccd images. Further studies concerning the influences of the ethanol on the discharge mechanisms in a rf APGD plasma with argon-ethanol mixture as the plasmaforming gas are necessary in future work. ACKNOWLEDGMENT This work has been supported by the project sponsored by the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry of China. The iccd images presented in this paper were taken using the iccd camera of the Plasma Lab, the Institute of Mechanics, Chinese Academy of Sciences. 1 A. Schütze, J. Y. Jeong, S. E. Babayan, J. Park, G. S. Selwyn, and R. F. Hicks, IEEE Trans. Plasma Sci. 26, M. Laroussi, IEEE Trans. Plasma Sci. 24, N. Puač, Z. L. Petrović, G. Malović, A.Ðorđević, S. Živković, Z. Giba, and D. Grubišić, J. Phys. D 39, J. Toshifuji, T. Katsumata, H. Takikawa, T. Sakakibara, and I. Shimizu, Surf. Coat. Technol. 171, T. C. Montie, K. Kelly-Wintenberg, and J. R. Roth, IEEE Trans. Plasma Sci. 28, M. Moisan, Z. Zakrzewski, R. Etemadi, and J. C. Rostaing, J. Appl. Phys. 83, M. Miclea, K. Kunze, U. Heitmann, S. Florek, J. Franzke, and K. Niemax, J. Phys. D 38, J. Park, I. Henins, H. W. Herrmann, G. S. Selwyn, J. Y. Jeong, R. F. Hicks, D. Shim, and C. S. Chang, Appl. Phys. Lett. 76, G. S. Selwyn, H. W. Herrmann, J. Park, and I. Henins, Contrib. Plasma Phys. 41, J. J. Shi and M. G. Kong, J. Appl. Phys. 97, J. J. Shi and M. G. Kong, IEEE Trans. Plasma Sci. 33, J. J. Shi and M. G. Kong, Phys. Rev. Lett. 96, J. Laimer, S. Haslinger, W. Meissl, J. Hell, and H. Störi, Vacuum 79, X. Yang, M. Moravej, G. R. Nowling, S. E. Babayan, J. Panelon, J. P. Chang, and R. F. Hicks, Plasma Sources Sci. Technol. 14, S. Y. Moon, J. K. Rhee, D. B. Kim, and W. Choe, Phys. Plasmas 13, J. Y. Jeong, S. E. Babayan, V. J. Tu, J. Park, I. Henins, R. F. Hicks, and G. S. Selwyn, Plasma Sources Sci. Technol. 7, J. Y. Jeong, S. E. Babayan, A. Schütze, V. J. Tu, J. Park, I. Henins, G. S. Selwyn, and R. F. Hicks, J. Vac. Sci. Technol. A 17, S. E. Babayan, J. Y. Jeong, V. J. Tu, J. Park, G. S. Selwyn, and R. F. Hicks, Plasma Sources Sci. Technol. 7, S. E. Babayan, J. Y. Jeong, A. Schütze, V. J. Tu, M. Moravej, G. S. Selwyn, and R. F. Hicks, Plasma Sources Sci. Technol. 10, G. R. Nowling, S. E. Babayan, V. Jankovic, and R. F. Hicks, Plasma Sources Sci. Technol. 11, H. W. Herrmann, I. Henins, J. Park, and G. S. Selwyn, Phys. Plasmas 6, V. J. Tu, J. Y. Jeong, A. Schütze, S. E. Babayan, G. Ding, G. S. Selwyn, and R. F. Hicks, J. Vac. Sci. Technol. A 18, M. Vleugels, G. Shama, X. T. Deng, E. Greenacre, T. Brocklehurst, and M. G. Kong, IEEE Trans. Plasma Sci. 33, S. Perni, X. T. Deng, G. Shama, and M. G. Kong, IEEE Trans. Plasma Sci. 34, K. A. Svitil, Discover 18, J. Park, I. Henins, H. W. Herrmann, and G. S. Selwyn, J. Appl. Phys. 89, J. Park, I. Henins, H. W. Herrmann, G. S. Selwyn, and R. F. Hicks, J. Appl. Phys. 89, J. J. Shi, X. T. Deng, R. Hall, J. D. Punnett, and M. G. Kong, J. Appl. Phys. 94, J. J. Shi and M. G. Kong, Appl. Phys. Lett. 87, J. Laimer, S. Haslinger, W. Meissl, J. Hell, and H. Störi, Proceedings of the 17th International Symposium on Plasma Chemistry, Toronto, Canada, 7 12 August 2005, edited by J. Mostaghimi, T. W. Coyle, V. A. Pershin, and H. R. Salimi Jazi, pp S. Wang, V. Schulz-von der Gathen, and H. F. Döbele, Appl. Phys. Lett. 83, J. Park and I. Henins, U. S. Patent No. US B1 21 June Y. P. Raizer, M. N. Shneider, and N. A. Yatsenko, Radio-Frequency Capacitive Discharges CRC, Boca Raton, FL, 1995, pp J. Glosík, J. Pavlík, M. Šícha, and M. Tichý, Czech. J. Phys., Sect. B 37, J. Muñoz, M. Pineda, M. Jiménez, I. Santiago, and M. D. Calzada, 32nd EPS Conference on Plasma Phys. Tarragona, Spain, 27 June 1 July, 2005, Vol. 29C, P S. Y. Moon, W. Choe, and B. K. Kang, Appl. Phys. Lett. 84, S. Y. Moon and W. Choe, Spectrochim. Acta, Part B 58, F. Tochikubo, Z. L. Petrović, S. Kakuta, N. Nakano, and T. Makabe, Jpn. J. Appl. Phys., Part 1 33,